Jennifer M. Mantooth. Digital Preservation Activities and Attitudes in American and Canadian Academic Libraries. A Master’s paper for the M.S. in I.S. degree. April, 2009. 59 pages. Advisor: Christopher Lee. Digital materials are highly used resources in academic libraries and absorb a significant cost in their creation. While digital materials greatly improve access to and usability of information, these materials must be actively maintained and preserved. Recent surveys, however, indicate that many libraries are not implementing adequate digital preservation techniques to insure their investments. This study surveyed American and Canadian institutions regarding their digital preservation processes and attitudes, with a specific focus on the comparison of processes and attitudes in institutions that have participated in a national digitization or digital preservation grant (grant receiving institutions) and non grant receiving institutions. The results suggest that grant receiving institutions are just as likely to implement migration, emulation, or digital archeology for digital image files as non-grant receiving institutions and that migration is the most widely implemented preservation strategy and is perceived as more effective than emulation or digital archeology by ARL institutions.

Headings:

Digital Preservation -- United States

Digital Preservation -- Canada

Academic Libraries -- United States

Academic Libraries -- Canada

Digital Preservation Activities and Attitudes in American and Canadian Academic Libraries

by Jennifer M. Mantooth

A Master’s paper submitted to the faculty of the School of Information and Library Science of the University of North Carolina at Chapel Hill

in partial fulfillment of the requirements for the degree of Master of Science in

Information Science.

Chapel Hill, North Carolina

April 2009

Approved by

_______________________________________

Christopher Lee

1 

 

Introduction

As the human experience continues to become increasingly intertwined with

technological tools, its documents likewise become increasingly dependent on the

surrounding technology. Today’s archivists are faced with the task of preserving digital

objects and the related technology required to make sense of them. However, Hedstrom

argues that it is not solely the archive’s concern. The unique aspects of digital objects

require early intervention for successful long-term preservation. Thus the traditional

preservation roles held by archivists are creeping across professional boundaries.

Government agencies, corporations, and research libraries all hold considerable amounts

of digital information that will need to be preserved into the future and it is the

organization’s responsibility to ensure preservation (Hedstrom 3).

A discussion of the nature of digital objects helps to explain their increasing

ubiquity in libraries as well as the unique challenges of digital preservation. Electronic

documents can improve accessibility to heavily used resources, allow users to manipulate

and annotate the information, as well as increase access points to underutilized

information. Certainly, these benefits play a role in libraries’ decisions to include digital

objects in their holdings. However, electronic documents have unique preservation needs

that cannot be fully met by the implementation of traditional preservation processes.

Terry Cook writes that unlike their paper-based predecessors, digital documents are not

eye-readable (Cook 301). Instead, digital objects are dependent on technology, which is

constantly evolving, for interpretation. Digital preservation is further complicated by the

2 

 

digital object’s reliance on contextual information and documentation (Ross 13).

Today’s libraries, however, are not facing digital preservation empty handed.

The archival community has discussed the challenge of long-term preservation

of electronic documents through the development of content preservation strategies:

migration, emulation, and digital archeology and has published numerous articles

discussing the perceived benefits and limitations of each. Librarians and archivists have

both been involved in the development of open access digital preservation tools such as

PRONOM, an online file format registry and the Reference Model for an Open Archival

Information System (OAIS), which provides a framework for long-term digital

preservation and access.

Wrubel writes that “Electronic resources constitute an increasingly significant

portion of library collections, both in usage and cost” (Wrubel 225). It seems then, that

it is in the best interest of libraries and archives to implement adequate digital

preservation policies to insure the continuing utility of their most valued resources, and

justify the expense of creating or obtaining them. However, there seems to be a large

spectrum of awareness and application of digital preservation techniques. While the

Koninklijke Bibliotheek (KB), for example, is actively participating in a robust digital

preservation strategy to preserve both content and access long term, there seems a large

majority of institutions that are not actively engaged in or perhaps even concerned with

digital preservation.

In an article published in 2006 attempting to inform institutions of the benefits

and processes of digitization, A. Bhatnagar expresses awareness that obsolescence is a

huge threat to digital materials and is careful to note the high financial cost of

3

 

digitization. However Bhatnagar fails to mention the cost of a long-term preservation

strategy to keep these objects readable, and thereby insure the institutions initial

investment in their digitization. In light of this research, it may come as little surprise

then, that a 2005 survey of the digital preservation practices of Museums, Libraries and

Archives in the United Kingdom revealed that while 23 out of 26 institutions surveyed

acknowledged digital holdings (Simpson 15), only 6 had a digital preservation policy in

place (Simpson 9).

A similar study was conducted in Australia in 2005. Pymm and Lloyd

interviewed library staff at four state libraries and the National Library of Australia

(NLA) on the acquisition, preservation, and accessibility of digital materials at the

library. Pymm and Lloyd state that their purpose was “to gain a broad overview of their

approach to handling their digital collections and of concerns they may have about

trying to ensure long-term accessibility to such collections” (169). The researchers

found that while the state libraries were aware that many of their digital holdings were

now obsolete, “no library reported any program to review holdings and establish any

migration/copying program while the technology may still be available” (Pymm 176).

Additionally, all state libraries reported that they lacked the resources for digital

preservation and were looking instead to the NLA to take the lead, or to partner with

other libraries and archives to preserve digital materials. (Pymm 170 -171).

Findings of other surveys, however, provide evidence of progress. A survey

conducted by Michele Cloonan and Shelby Sanett indicates that a significant portion of

libraries and archives are actively involved in digital preservation activities. Cloonan

and Sanett conducted a survey of thirteen institutions and archival projects in the US in

 

4

 

2002. Each of the institutions surveyed employed a member of the InterPARES

Preservation Task Force. The study reports that 10 of 13 institutions surveyed

(approximately 75%) were involved in a digital preservation strategy. While the survey

conducted in the United Kingdom and Australia sampled institutions from the entire

population of libraries, archives, and museums in the country, the American study

sampled institutions from the International Research on Permanent Authentic Records

Electronic Systems (InterPARES) project. Kyong-Ho Lee et al. write that the

InterPARES project is “a multinational research initiative…collaborating to develop the

theoretical knowledge and methodology required for the permanent preservation of

authentic records” (101). The fact that all of the institutions surveyed in the American

study were institutions with demonstrated commitment to digital curation may account

for the high percentage of institutions involved in a digital preservation strategy.

It seems that many institutions are not actively engaged in digital preservation

and a survey of the literature may provide insight into one of the reasons why. The

literature does not universally agree on various aspects of the effectiveness of digital

preservation techniques, such as the cost of migration and emulation. Archival

institutions must continue to support current work processes as they investigate

potential preservation strategies, the contradictory literature may be viewed as a

hindrance to the individual looking for a quick answer. It seems that institutions actively

engaging in grant projects such as interPARES are more likely to display digital

preservation activity. The realization of this implication appears to exist as a void in the

current literature.

 

5

 

Purpose:

This research surveyed American and Canadian research institutions regarding

their digital preservation processes and attitudes. The main focus of the study was to

explore preservation activities and perspectives at these institutions. An additional focus

was the comparison of processes and attitudes in institutions that have participated in a

national digitization or digital preservation grant (grant receiving institutions) and non

grant receiving institutions. The study also compared the attitudes towards digital

preservation techniques with the literature.

Research Questions:

1. What are the general trends in digital preservation practices and attitudes at

American and Canadian research institutions?

2. Are grant receiving institutions more likely to be actively involved in digital

preservation activities?

3. Is there a relationship between the presence of digital preservation activities at

the institution and the attitudes about the techniques used?

Literature Review

Overall, the literature offers valuable, high level discussions of digital

preservation techniques, what perhaps is missing is an appreciation of the individual

archival institution’s perceptions of the preservation options available.

Much of the discussion on digital preservation techniques is published in online

trade journals such as D-Lib Magazine and RLG Digi-News (now defunct). When

possible, case studies are used to illustrate the theoretical benefits and/or perceived

shortcomings of the approach.

 

6

 

Reliance on Standards

Many digital preservation techniques, including migration and emulation

combine reliance on standards with additional preservation activities. Rothenberg writes

that “converting digital documents into standard forms, and migrating to new standards

if necessary, may be a useful interim approach while a true long-term solution is being

developed”. The Library of Congress defends the importance of formal technical

standards for the successful creation and maintenance of digital objects, and especially

digital images, as evidenced by a compendium of digital image standards published by

the Library of Congress in 2004 (Peterson). Rothenberg, however, cautions that one of

the limitations of relying on standards is that formats “are still evolving so rapidly that it

is unrealistic to expect definitive standards for any of these forms to emerge in the near

future” (Rothenberg). Yet even today, the literature indicates that formats continue to

evolve and cloud the standards issue, as illustrated by the debate over the defacto

standard for digital image archiving.

In 2008, Paolo Buonora and Franco Liberati write that digital libraries prefer

JPEG 2000 to TIFF, and that this format has in fact replaced TIFF as the defacto

standard for digital image archiving. Yet, while supporting JPEG 2000 as a viable

preservation format, Robert Buckley cautions that “ JPEG 2000 is lossy and

irreversible, [although] the losses are not visually noticeable” (26). Andreas Stanescu

writes that lossy compression is “an impediment to preservation” and advises the use of

the TIFF file format over any lossy format (74). The controversy over the

appropriateness of JPEG 2000 to replace TIFF as the defacto standard for digital images

offers evidence that Rothenberg’s claim, issued nearly a decade ago, still holds true.

 

7

 

Rothenberg indicates that another limitation of reliance on standards as a digital

preservation technique stems from the issue of forcing incompatible document types, or

paradigms, into the ‘standardized format’. Rothenberg uses the term paradigm to

indicate the vastly different methods of display, information storage, and functionality

of technical classes of documents such as databases, spreadsheets, e-mail attachments,

and images. Rothenberg writes that reliance on standards necessitates paradigm shifts

which may lead to a great loss of information as the original format is forced into a

standardized form. Rothenberg indicates that when paradigms are drastically different,

such as relational databases and object oriented databases, the translation of one to

another loses so much information that the result risks becoming meaningless.

Digital Archeology

Computer museums are the critical force behind digital archeology. Digital

archeology, also known as retrospective archiving (retro-archiving), refers to the

process of retrieving information from obsolete media. One of the benefits of digital

archeology is that the artifact can stay in its original state or paradigm and maintain its

original functionality (Rothenburg). Digital archeology is not considered a responsible

preservation technique on its own, and is generally combined with either migration or

conversion to a standardized format. A limitation of this approach is the expense.

Rothenberg writes that digital archeology becomes highly labor-intensive because the

computer museums need to maintain hardware and software drivers to enable the

obsolete machines to read the new storage media devices. Kyong-Ho Lee et al., add that

computer museums and by extension digital archeology may be impossible to support

over the long term due to equipment deterioration and lack of vendor support (95).

 

8

 

Despite their limitations, a study conducted in 2007 indicates that digital

archeology is becoming an increasingly important preservation technique. In 2007,

Balkestein and Tjalsma report the results of the Archiving Digital Academic Heritage

project, a three-year project focused on using digital archeology to provide access to

digital research datasets on obsolete media at the Meertens Institute in the Netherlands

(95). Even though the data sets were on media that was ten to thirty years old, with the

help of the Computer Museum of the University of Amsterdam, the project was able to

read approximately 97% of the obsolete formats, only 52 of 1599 were unreadable

(Balkestein and Tjalsma 96). While the project was technically successful, Balkestein

and Tjalsma indicate that one of the greatest limitations of digital archeology is the loss

of contextual information necessary to give meaning to the raw data. Without this

contextual information the recovered files had much less value (96). The literature

indicates that libraries and archives will continue to be heavily dependent on

incorporating digital archeology into their digital preservation activities. An institutional

report from the National Library of Australia suggests that without the external help of a

computer museum approximately 1/3 of obsolete media (in this case up to 12 years old)

will be unreadable (Colin).

 Migration

In Network Notes #60, published by the National Library of Canada, Alison

Bullock defines migration as a process that “covers a range of activities to periodically

copy, convert or transfer digital information from one generation of technology to

subsequent ones.” Oltmans indicates that a benefit of migration is that the objects will

usually be in a format that is easily accessible across various computing platforms.

 

9

 

Rothenberg offers a scathing review of migration, which reinforces his role as what

Granger deems “the staunchest supporter of emulation”. Rothenberg lists among the

limitations of migration that it is error-prone, expensive, and difficult to automate.

Much of the literature agrees that one of the largest weaknesses of migration is

the potential for data loss. Oltmans cautions that a single conversion from one form to

another may result in loss of data. Mellor discusses the potential for a process of

compounding migration errors from one migration to another throughout future

migrations and resulting in an object that may have little resemblance to the original

(517). Bullock agrees that successive migrations may have the potential to lead to

unacceptable data loss.

Rothenberg indicates that much of the expense of migration resides in the

inability to safely automate the process. He writes that each different document type

requires a unique solution, independent technology forecasting to determine when

migration is necessary, and variable migration schedules, all of which require

independent handling and cannot be responsibly applied to different document types

(Rothenberg). Mellor attributes the high cost of migration to the need to continually

produce migration tools and then apply them to a very large number of data objects.

Granger argues against the attribution of ‘expensive’ as a shortcoming of migration, he

asks “expensive compared with what?” and writes that migration has been the “only

serious candidate thus far for preservation of large scale archives.”

Despite the perceived shortcomings of migration, Kranch, Bullock, Granger, and

Rothenberg agree that migration is the most common digital preservation strategy

among archival institutions. The OAIS Reference Model cautions that migrating digital

 

10

 

content is time consuming, expensive, and may expose the repository to a “greatly

increased” probability of loss of information (Reference 91) Despite the caution, the

OAIS model seems to prefer migration over emulation techniques which the authors

write are not mature enough for “significant comment” (101) The model contains a

detailed list of potential motivations for migration (90 – 91), offers a migration

perspective for the model (92) and discusses various migration types (93 – 98).

In an effort to allay the costs associated with constant migrations to the

continually evolving popular formats, Mellor proposes a variation on traditional

migration: migration on request. Unlike traditional migration, which hinges on

processes that transform the copy of data that is stored within the repository, migration

on request preserves the original digital object and migrates the object, using an

algorithm based on its format, only when the object is requested for viewing (Mellor

518). Based on the results of a migration on request tool, developed by Mellor and a

team of researchers, Mellor indicates that “over a short to medium term period

migration on request should offer major cost savings in comparison to a traditional

migration strategy” (526).

Emulation

While there are significant amounts of literature dealing with emulation, the

literature seems to confuse an already technically complex issue and there is argument

over the perceived benefits and limitations of emulation. This confusion may offer

insight into the overwhelming preference for migration in archival institutions, as well

as the tendency of archival institutions with demonstrated commitment to digital

 

11

 

curation, such as participation in the InterPARES programs, to be more engaged in

digital preservation.

Oltmans defines emulation as a process which “preserves the authentic

document and provides the user with a tool that enables ‘old’ software and ‘old’ viewer

programs to render this original document”. In Oltman’s definition of emulation, the

document and the application software are original, and the emulation imitates either

the operating system or the underlying hardware. However, Bullock defines emulation

as the process of “creating new software that mimics the operations of older hardware

of software” While Bullock’s definition may be more encompassing than Oltmans (she

includes the potential for all three types of emulation: software, operating system, and

hardware), it indicates the difficulty of making sense of the literature. One may suspect

that each type of emulation carries its own challenges and positive affordances.

However, when emulation is dealt with in the literature, it generally is not prefaced by

an identification of type.

Another confusing factor is the issue of the ability to test an emulator

beforehand. Again, the literature is not consistent with respect to this aspect. While

Rothenberg writes that one of the great strengths of emulation is the ability to test the

accuracy of the emulator in advance, Oltmans indicates that one of the weaknesses of

emulation is that one cannot know for certain that they will always work on future

computer platforms. The reader may assume that the difference may depend on the type

of emulation and the length of time projected, but it still does little to illuminate the

issue.

 

12

 

One thing that the literature consistently agrees upon is the ability for emulation

to preserve the original ‘look and feel’ of the digital object. (Oltmans, Bullock,

Rothenberg, and Granger). However Margaret Hedstrom and other researchers

conducted an experiment in 2006 in which thirty subjects evaluated the performance of

a twenty-year-old computer game in its native environment, in an emulated

environment, and migrated to a Windows environment (168). The team reports that they

“found no evidence that emulation is better for preserving original ‘look and feel’ than

migration” (Hedstrom et al. 171). In fact, the study reports that few subjects actually

lamented the loss of the original ‘look and feel’ of the computer game, most instead

preferring the improved functionality of the migrated and emulated versions (Hedstrom

et al. 171). It may be interesting to point out that in this experiment the display of the

original was not copied exactly in the emulated version which imposed a bar across the

screen (Hedstrom et al. 170).

A cost comparison of migration and emulation by Erik Oltmans and Nanda Kol

of the Koninklijke Bibliotheek (KB) in the Netherlands provides contrary evidence to

the generally accepted claim that emulation is more expensive than migration. The KB

has collaborated with IBM to develop and implement the e-Depot, a system that

preserves both the content of and accessibility to digital publications (Oltmans and

Kol). The validity of the research team’s findings may be supported by the evidence

that the KB is acclaimed worldwide for its expertise in digital preservation and in fact

has gained the confidence of 6 major publishers including Elsevier Science and

Springer to archive their electronic publications (Oltmans and Kol). Generating cost

 

13

 

comparisons from literature and personal insights, Oltmans and Kol conclude that, in

the long term, emulation is more cost efficient than migration.

Oltmans and Kol base their conclusion on the issue that migration entails a

steady stream of resources to continually monitor and convert objects; emulation on the

other hand requires high initial costs (Oltmans and Kol). A significant virtue of

emulation, in the realm of costs, is that once it is created an emulation tool can be

applied to an entire collection of objects as opposed to a repetitive single object at a

time approach as in migration (Oltmans and Kol). Oltmans and Kol argue that it is this

property of emulation that makes it more cost effective than migration for long-term

preservation. According to a cost projection for preserving 1 million objects for 50

years, Oltman and Kol report that the cost of migration is 79% more than the cost of

emulation. When the authors increase the number of objects in the archive to 5 million,

after 50 years the projected cost of migration is more than 200% over the cost of

emulation. Although perhaps the bulk of the costs for emulation reside in the creation of

the emulator itself, Oltmans and Kol advise that these costs can be distributed and

thereby lessened if digital repositories globally share their emulators with each other.

Other Techniques:

While much of the literature discusses well-accepted digital preservation

techniques, the literature does contain processes that have not made it into the

mainstream. Reliance on print outs and digital tablets are examples of such techniques.

Reliance on Print Outs

Jeff Rothenberg defines reliance on hard copy as to print digital documents onto

paper and then follow traditional archival processes to preserve that paper. A benefit to

 

14

 

this technique would be that archives and libraries could use traditional physical

preservation techniques that have been used to preserve paper based archival holdings.

However, the translation of a digital born object to an analog document has the potential

to lose many if not most of the positive affordances innate to that object, and also a

potentially great deal of information as well. Margaret Hedstrom and Christopher Lee

define significant properties as “those properties of digital objects that affect their

quality, usability, rendering, and behavior” (218). Rothenberg notes that a significant

limitation of reliance on print outs is that it forfeits the benefits of machine readability

and obliterates the interactive functionality within and between digital documents.

Digital Tablets

In 1998, Douglas Kranch published an article that champions digital tablets as a

superior and inevitable alternative to migration and emulation. Kranch’s digital tablets

would preserve the hardware and software required to access the digital objects, as well

as the object itself. In addition, the tablets would contain a large amount of read only

storage (10s or hundreds of terabytes), reside in a protective plastic encasing, and run

off their own power supply. Kranch indicates that a benefit to digital tables would be

the ability to preserve the original document and its software in order to preserve the

functionality and “feel” of the document. Despite the initial attention, digital tablets are

not mentioned as a viable preservation strategy in later literature.

Stepping back

While most of the literature agrees that reliance on print outs, reliance on

standards, and digital archeology are not valid preservation strategies on their own,

there continues to be discussion over the comparative strengths and weaknesses of

 

15

 

migration and emulation. Much of the literature advises the individual to step back from

the quest for the “silver bullet” preservation strategy and examine the issue as a whole.

Contributing factors

While technological strategies no doubt have an impact on the success of a

preservation program, the literature suggests that other factors come into play as well.

Rothenberg indicates that although “a sound technical approach must form the

foundation on which everything else rests”, administrative, organizational, and policy

issues also impact the success of a preservation strategy. In contrast to Rothenberg,

Nancy McGovern advocates that digital preservation be supported by three main

components, which can be seen as three legs of a stool: organization, technology, and

resources. She advises that “a balanced three-legged stool … will be more effective in

establishing a sustainable digital preservation program than a technology pogo stick”

(McGovern).

A case study conducted by Howard Besser in 2006 attributes the success of the

Preserving Digital Public Television project, a highly collaborative digital preservation

project engaging 3 large, high profile public television stations and New York

University, in part to the organizational commitment of the parties involved. Besser

writes “from early on in the project there was active high-level commitment and support

from most of the players (at the vice-president level from the two stations, at the dean’s

level from the university).” (223).

Dual approach

While authors exist on either end of the spectrum, much of the literature

advocates a combination of digital preservation techniques. Both Rothenberg and

 

16

 

Bullock support the use of standards in maintaining cost efficiency. Mellor advises

institutions to implement migration on request along with emulation (526).

Different approaches support different needs

The literature encourages the archival institution to understand the unique

characteristics of their own collections when choosing preservation techniques. Oltmans

and Kol discuss the importance of what Hedstrom and Lee term as the significant

properties of a digital document when selecting a preservation strategy. If an

organization has a commitment to preserve the authentic form of a document, or if the

original look and feel is deemed a significant property of the object, Oltmans and Kol

propose emulation over migration. However, if an archive anticipates that future users

will benefit from viewing an object according to the standards and functionalities they

have come to expect from technology migration is a better strategy (Oltmans and Kol).

Methodology

This research gathered information about the digital preservation activities of

academic libraries and the perspectives of key members of these institutions regarding

well-known digital preservation techniques. Academic libraries will be the units of

analysis for this study. In order to study the practices and attitudes of institutions, this

study surveyed individuals who acted as informants for the institutions. Surveys are

considered an effective tool for measuring attitudes (Babbie 244). Informants provided

information regarding the digital preservation practices of the institutions and the

institutional perspective on digital preservation techniques. Informants were recruited

through email to participate in a web-based study. This enabled the study to examine

institutions without the limitations of geography. The survey targeted individuals with

 

17

 

a fair amount of authority in the institution, based on the assumption that they are aware

of digital preservation activities taking place at the institution. Additionally, the digital

preservation perspectives of these individuals may have played a larger role in the

institutional perspective. Most of the survey questions were close-ended, enabling

respondents to chose an answer from a list. A benefit to close-ended questions is that

they “provide uniformity of responses and are more easily processed than open-ended

questions” (Babbie 246).Care was taken to ensure that the survey questions were

concise and addressed only a single topic.

Sample, Population, and Sampling Technique

This study recruited contacts from the Association of Research Libraries (ARL)

member list from the ARL website. ARL is a not-for-profit organization of research

libraries in the United States and Canada. There are currently 123 members of the ARL.

Subject Identification and Recruitment:

All the academic libraries surveyed came from a variety of American and

Canadian institutions with digital collections. 17 of the 26 respondents, who indicated

their participation in grants, had received a national digitization and/or digital

preservation grant, 7 had not received one, and 2 did not know.

Libraries do not generally follow the same organizational hierarchy, and job

roles vary widely across institutions. However, a list of head administrators of each

library and their emails was constructed by the author. These individuals were emailed a

request to forward the survey to the individual or individuals at their institution who are

most capable of providing answers regarding the institution’s digital preservation

 

18

 

practices. One week after the initial email was sent, a follow up email was sent to

encourage response.

Variables:

Previous participation in a national digitization or digital preservation grant was

a quasi-independent variable in this study. Digital preservation practices and attitudes

was analyzed independently within these groups and then against each other.

Engagement in migration, emulation, and digital archeology also acted as a quasi-

independent variable in later data analysis.

Data Collection Instrument

Data was collected by means of an online survey. The survey was divided into

three parts. The first part provided identifying characteristics of the informant and the

institution. The following five questions composed the first part of the survey:

1. Please identify the location of your institution (U.S. Canada Other)

2. Please indicate your job title.

3. Please provide the name of your institution. (Note: this information is

used to identify duplicate responses and will not be included in any

derived report).

4. Within the past 10 years, has the institution been involved in a national

digitization and/or digital preservation grant?

(Y N Applied but didn’t receive funding Don't know)

Institutional names were replaced with a generic identifier (A,A1, etc) at the

close of the survey in order to maintain the confidentiality of institutions in the raw data

and in the analysis.

 

19

 

The fifth question asked respondents to describe the contents and size of their

institution’s digital collections.

The second part of the survey provided information regarding the institutions

digital preservation activities.

In the following two main and sub questions, the informant was asked to provide

information regarding their digital text and image files. These questions were open-

ended to encourage honesty. The large number of potential file formats precluded the

use of a list of possible answers, and at any rate such a list arguably would have never

included all possible formats. These questions provided insight into the institution’s use

of file standards in their preservation activities. Additionally, they helped indicate if the

institution was following best practice regarding storage media and file refreshing

timelines. The final function of these questions was to verify the truthfulness of the next

question. That is, if an informant responded that the preservation copies of their image

files were mostly on the original 15 year old CDs and later indicated that the image files

had been migrated, the researcher could note the inconsistency.

5. Please provide the following information for digital TEXT files:

a. What file format(s) do you use for the digital master copies of your

digital text files?

b. On what physical media do you store your digital text collections?

c. Are you storing your digital text collections on physical media that are

different from those to which you originally stored them?

d. What is the age range of these materials?

6. Please provide the following information for digital IMAGE files:

 

20

 

a. What file format(s) do you use for the digital master copies of your

digital image files?

b. On what physical media do you store your digital image collections?

c. Are you storing your digital text collections on physical media that are

different from those to which you originally stored them?

d. What is the age range of these materials?

The next questions were arranged in a table to enable ease of response. Survey

participants were asked to indicate institutional participation in migration of image files

and of text files, emulation of image files and of text files, and digital archeology of

image files and of text files. Possible answers were Yes, No, or Don’t Know. Text and

image files were handled separately in order to avoid barrel questions. Digital

archaeology was defined in the questions as ‘extracting information from obsolete

media – yourself or by hiring a vendor to do the extraction’ in order to clarify the

technique to participants.

7. Please indicate whether your institution has used migration in the

management of your digital image files

8. Please indicate whether your institution has used emulation in the

management of your digital image files

9. Please indicate whether your institution has used data archeology

(extracting information from obsolete media – yourself or by hiring a

vendor to do the extraction) in the management of your digital image

files

 

21

 

10. Please indicate whether your institution has used migration in the

management of your digital text files

11. Please indicate whether your institution has used emulation in the

management of your digital text files

12. Please indicate whether your institution has used data

archeology(extracting information from obsolete media – yourself or by

hiring a vendor to do the extraction) in the management of your digital

text files

The third part of the survey provided information regarding the informant’s

perspective on migration, emulation, and digital archeology. The informant was asked

to provide his or her personal reaction to the statements that followed. As stated earlier,

informants were selected based on their knowledge of digital preservation practices in

the institution. Even though the institution is the unit of analysis for the study, in this

case, the informant was the closest voice of the institution available. Responses to these

questions were based on a five-point Likert scale (Strongly Agree, Agree, Neither

Agree nor Disagree, Disagree, or Strongly Disagree). This format is generally quick to

complete, and can increase the ability to compare responses across questions (Babbie

254). All five statements about a given preservation technique as well as the

corresponding Likert scale were grouped to be visible on one computer screen for each

preservation technique. These questions provided comparable information about the

attitudes of the informant regarding three major digital preservation techniques and

enabled the researcher to create descriptive statistics of digital preservation

 

22

 

perspectives. The instruction to the informant and the statements for each technique are

listed below.

Please indicate your level of agreement with the following statements:

Migration:

is labor intensive

has a high financial cost

is an effective preservation strategy for digital text

is an effective preservation strategy for digital images

is widely implemented

Emulation:

is labor intensive

has a high financial cost

is an effective preservation strategy for digital text

is an effective preservation strategy for digital images

is widely implemented

Digital Archeology (extracting information from obsolete media – yourself or by

hiring a vendor to do the extraction):

is labor intensive

has a high financial cost

is an effective preservation strategy for digital text

is an effective preservation strategy for digital images

is widely implemented

Advantages and Disadvantages of Method

 

23

 

Babbie indicates that the strengths of surveys include the ability to describe

characteristics in a large population and the ease of information gathering from a large

sample size (276). The weaknesses of surveys include the inability to delve deeply into

complex issues in the same manner that observational methods can and the inability to

modify the tool after the study is underway (Babbie 276).

Results

Of the 123 ARL institutions, the researcher was only able to discover 113 top

library administrators and their email addresses. 113 heads of the libraries at ARL

institutions were recruited to participate in the survey and 43 surveys were started.

Surveys in which the respondent exited immediately after clicking “accept” on the

consent form, which was the first question of the survey, were removed from the

resulting dataset. This removed 7 surveys from the resulting dataset, and indicates that

the true response rate is 34/113 or approximately 30%.

Duplicate responses

The survey respondents are informants into an ARL institution’s digital

preservation activities and their own attitudes. Due to the nature of survey distribution,

the researcher anticipated that she may receive several responses from different

informants at the institution. However, the unit of analysis for this study is the ARL

institution, so it was deemed important that no institution be over-represented in the

resulting dataset. In order to ensure that one institution was not overly represented in the

resulting data, the survey results were examined in order to identify duplicates. IP

addresses, institution names, and the job titles of respondents were examined and 5

institutions appeared to have duplicate survey responses.

 

24

 

Three institutions had two surveys taken by the same individual. Based on the IP

address, the institution name, the respondent’s job title, and the respondent’s note in one

of the partial surveys, the researcher was able to combine two partial surveys taken by

the same individual at one institution. The researcher was contacted by one participant

who completed two surveys for the same institution and was informed that the most

recent survey was most accurate, as a result she removed the older duplicate from the

resulting dataset. In the same manner and in order to minimize the effects of

duplication, the less complete survey for the third institution was removed.

The fourth and fifth institutions both had two surveys taken by two different

individuals. In both cases, one survey was minimally complete and the other was

complete. The responses were identical except for the job title. As a result, the

researcher removed the partially complete survey from the resulting dataset. The

resulting, non duplicate dataset consists of 31 surveys.

Respondent background

28 known institutions were represented in the 31 responding, non duplicate

surveys. 3 respondents did not identify the name of their institution. Of the 31

institutions represented, 27 were in the United States and 4 were in Canada.

Of the 31 survey responses, 3 respondents did not identify their job title. Of the

28 respondents that provided their job title, 20 (over 70%) held positions of

considerable authority in the library. Three respondents identified themselves as the

head of the library (Director of Libraries, University Librarian, Dean of Libraries), one

was the Deputy Director, and 16 respondents held job titles that indicate that the

respondent was the director or head of a department within the library.

 

25

 

Specialization

Many surveys were completed by individuals that held job titles that indicate

specialization in digital technologies. 18 of the 28 respondents had the term “digital” or

“technology” in their job titles. 6 of 28 respondents had the term “preservation” in their

job title. Four of the 6 respondents with the term “preservation” in their job title also

contained the term “digital” Only 2 individuals surveyed held job titles that contained

any variation on the term “archivist.”

Participation in Grants

Five respondents did not answer this question.17 of 26 respondents (65%)

indicated that their institution had been involved in a national digitization and/or digital

preservation grant. Two of 26 (8%) indicated that a grant had been applied for but not

funded, and five of the 26 respondents (19%) for this question indicated that their

institution has not been involved in a national digitization and/or digital preservation

grant. Two respondents did not know if the institution had been involved in a

digitization grant in the past 10 years.

Digital Collections

Respondents were asked to describe the contents and size of their institution’s

digital collections. 25 informants responded to this question. Fourteen respondents

indicated the size of the digital collections only in terms of number of items. Five

respondents reported the size only in terms of Terabytes, Gigabytes, or Megabytes. Four

gave both item counts and approximate number of bytes to store the materials. Two

provided no indication of size in either number of items or number of bytes.

 

26

 

The number of items reported in an institution’s digital collections varied in

precision from ‘11,688’ to ‘several tens of thousands.’ They varied in size from one

institution that reported ‘approximately 2,200 images and one serial title’ to approx.

2,405,847 text, images, video, and learning objects. Of the 15 respondents who

provided a specific number of items, the mean number of items reported is 233,080.

The median number of items reported is 22,147.

Nine respondents provided the size of their digital collections in terms of

number of bytes. The answers ranged from 52 gigabytes to ‘no smaller than 40

terabytes’. The mean size of digital collections reported is 15 terabytes. The median is

5 terabytes.

Three respondents indicated that they anticipate their digital collections to

increase in size in the near future.

Collection contents varied. 24 of the 25 respondents indicated some combination

of text, image, audio, and video collections. One respondent provided no indication of

their collection contents. Five respondents indicate that their digital collections include

theses, dissertations, and scholarly publishing texts. Several respondents list websites

(2) and statistical datasets (4) among their digital holdings.

Digital Text Files

Respondents were asked to indicate the file formats used for the digital master

copies of their digital text files. There were 25 responses listing various file formats.

Among these responses there is a surprising presence of image based file formats. This

indicates that respondents were including scanned page images in this category, rather

than limiting their responses to born-digital text files. The text file formats mentioned

 

27

 

by respondents include PDF, TXT/ASCII and RTF. Figure 1 illustrates the distribution

of file formats in the responses.

Figure 1: Distribution of File Formats for Digital Text Files

All respondents reported that their digital text collections are stored on some

variation of what may be labeled large networked mass storage devices. Four

respondents indicate that CDs and/or DVDs are used in addition to servers. One

institution reports that while they do store some items on a server, most of their

purchased collections are housed and accessed from outside of the institution.

13 informants indicate that their institutions are storing their digital text collections on

physical media that is different from the original storage media. Three report that the

media is not different. 2 indicate that some media has been refreshed. 1 did not answer.

2 did not answer the question directly but indicated that backup copies are on DVDs or

CD-ROMs. Three did not know or wrote N/A.

The age range of the digital files varied from less than 1 year to up to 20 years

old. The distribution of files by age can be viewed in figure 2.

 

28

 

Figure 2: Distribution of the age range of digital text files at surveyed institutions.

Digital Image Files

Respondents were asked to identify the file formats used for the digital master

copies of their digital image files. Most respondents indicated formats for single

images, while one included file formats for documents that may comprise of scanned

text with embedded images (PDF, PDF/A). There were 26 responses, some of which

listed a variety of formats. 20 responses indicated only TIFF. Three responses listed

TIFF along with JPEG and/or JPEG2000. One response listed TIFF, PDF, and PDF/A.

Figure 3 illustrates the distribution of file formats in the responses.

 

29

 

 

Figure 3: Distribution of File Formats for Digital Image Files

The physical media used to store digital image collections match with the

physical media reported to store digital text collections (large networked mass storage

devices) for 25 of the 26 institutions that responded to this question. One respondent

indicates that while all the digital text files at his/her institution resides on an ‘archival

server’, some image files are on CD and the rest are stored on an ‘archival server’. The

respondent did not provide clarification on the term ‘archival server’.

The distribution of institutions that store their digital image collections on media

that is different from the original is similar to the counts for digital text collections. 15

respondents indicated yes. 5 indicated no. 3 indicated that some media has been

refreshed. 2 did not know or wrote N/A.

The age range of the digital files varied from less than 1 year to up to 20 years

old. The distribution of files by age can be viewed in figure 4.

 

30

 

Figure 4: Distribution of the age range of digital image files at surveyed institutions.

Digital Preservation Practices at ARL Institutions

Respondents were asked to indicate their institution’s use of three digital

preservation techniques (Migration, Emulation, and Data Archaeology) in the

management of its digital image and text files.

24 individuals provided information regarding the institution’s use of migration.

25 respondents answered the question about the institution’s use of emulation and data

archeology.

Participation in migration, emulation, and data archeology was sub-grouped

based on institutions that have received a grant in the past 10 years and those that have

not. The percentage responses are similar for both groups with slight differences.

Statistical tests were done to determine the statistical significance of these results. There

was no significant relationship found between receiving or not receiving a grant and

participating in migration or emulation of image files (χ2 =.8, df = 1, p = .37). There

was no significant relationship found between receiving or not receiving a grant and

participating in digital archeology or migration of image files, (χ2 =.02, df = 1, p = .88).

 

31

 

And there was no significant relationship found between receiving or not receiving a

grant and participating in digital archeology or emulation of digital image files (χ2 =.89,

df = 1, p = .35). There was not enough data for text files to perform statistical analysis.

Therefore, statistically, non-grant institutions are just as likely to participate in

migration, emulation, or digital archeology for their digital image files as grant

institutions. The percentage differences viewed in figure 5, below, are not statistically

significant.

Figure 5: Percent participation in migration, emulation, and data archeology for grant and non-grant

receiving institutions. The chart on the left shows the percentages for image files, the chart on the right

shows percentages for text files.

The standard deviation for grant receiving institutions that participated in

migration for their digital image and text files is .62 and .68, respectively. The standard

deviation for non-grant receiving institutions that participated in migration for their

digital image and text files is .71 and .64 respectively. The standard deviation for grant

receiving institutions that participated in emulation for their digital image and text files

is .50 and .49, respectively. The standard deviation for non-grant receiving institutions

that participated in emulation for their digital image and text files is 0 and .35

respectively. The standard deviation for grant receiving institutions that participated in

 

32

 

digital archeology for their digital image and text files is .53 and .50, respectively. The

standard deviation for non-grant receiving institutions that participated in digital

archeology for their digital image and text files is .64 and .53 respectively.

Due to the strong similarities between the use of migration, emulation, and data

archeology between grant receiving institutions and non-grant receiving institutions, the

results of the survey will be discussed in terms of the overall findings, as both groups of

institutions follow the same trends as the group to which both belong.

As illustrated in figure 6, more institutions report using migration, emulation,

and data archaeology on their digital image files than on their text files. Additionally,

more institutions are using migration than either data archaeology or emulation.

Emulation is the least used preservation technique.

Figure 6: Distribution of the use of migration, emulation, and data archeology

Figure 7 illustrates the relationship between the counts of institutions engaged in

migration, emulation, and data archeology of digital image files and the counts for

digital text files. Approximately 42% of institutions surveyed have used migration in

the management of their digital images and 33% have migrated digital text. 24% of

respondents indicated that their institutions have used data archeology on their digital

image files, while 12% have used data archeology on digital text files. The surveys

 

33

 

indicate that much fewer institutions are engaged in emulation. 8% of respondents

indicate that their institutions have used emulation for digital images and 4% indicate

that they have used emulation for digital text.

 

Figure 7: Counts of Instituons engaged in migration, emulation, and data archeology for their digital

images and digital text files

In addition, institutions are more likely not to use migration, emulation, or

digital archeology than to use them, regardless of file type. Figure 8 illustrates that all

counts of institutions using migration, emulation, or data archeology are lower than the

count of institutions that are not engaged in them.

 

34

 

Figure 8: Use of digital preservation techniques by type of file.

Digital Preservation Perceptions at ARL Institutions

Respondents were asked to indicate their level of agreement with statements

issued about different digital preservation techniques. 24 individuals provided responses

to each Likert scale question.

Perceptions on the Amount of Labor Expended in Migration, Emulation, and Data

Archeology

16 informants (67%) agreed or strongly agreed that both migration and data

archeology are labor intensive. Informants, however, felt more strongly about the labor

involved in data archeology than that of migration. Data archeology had a 12% higher

response for strong agreement with the statement that data archeology is labor intensive

than those who strongly agreed that migration is labor intensive.

A substantial percentage of informants neither agreed nor disagreed with the

statement. Over half of individuals surveyed (54%) neither agreed nor disagreed that

emulation is labor intensive, while 33% and 29% neither agreed nor disagreed that

migration or data archeology, respectively, are labor intensive. Figure 9 illustrates the

 

35

 

exact counts of survey responses to the statement that migration, emulation, and data

archeology (respectively) are labor intensive.

Figure 9: Institutional perceptions of the statement that migration, emulation, and data archeology

(respectively) is labor intensive, by institution count.

Perceptions of the Financial Expense of Migration, Emulation, and Data Archeology

The survey results indicate that data archaeology is considered most expensive

and emulation least expensive. Most respondents strongly agreed, agreed, or neither

agreed nor disagreed with the statement of financial expense. Only one respondent

disagreed that migration is financially expensive and no respondents disagreed that

emulation and digital archeology are financially expensive.

67% of participants agreed or strongly agreed that data archeology is financially

expensive. Respondents perceived migration to be more expensive than emulation. 54%

agreed or strongly agreed that migration is expensive and 46% believed the same for

emulation. Respondents seem to be more confident of the expense of data archeology

than the costs of emluation or migration. 8 respondents strongly agree that data

 

36

 

archeology is financially expensive while 4 respondents strongly agree that migration

and emulation are expensive. Figure 8 illustrates the exact counts of survey responses

to the statement that migration, emulation, and data archeology (respectively) are

financially expensive.

A substantial percentage of those suveyed neither agreed nor disagreed that

migration, emulation, and data archeology are financially expensive. Over half of the

responses neither agreed nor disagreed that emulation is financially expensive, 41 %

felt the same way about migration, and 33% indicated the same response for data

archeology.

Figure 10: Institutional perceptions of the statement that migration, emulation, and data archeology

(respectively) is financially expensive, by institution count.

Perceptions of the Effectiveness of Migration, Emulation, and Data Archeology as a

Digital Preservation Technique

There was very little distinction between the perceptions of the effectiveness of

digital preservation techniques for text files versus those for image files. In fact, the

 

37

 

responses given to the statements regarding the effectiveness of migration, emulation,

and data archeology for text files and the same for image files are either identical or

only off by one as may be viewed in figure 11.

Figure 11: The nearly identical responses given to the statements that migration, emulation, and data

archeology (respectively) is an effective preservation strategy for digital images and an effective

preservation strategy for digital text.

Few respondants indicated that they felt strongly about the effectiveness of

migration, emulation, or data archeology as a digital preservation technique and many

of those surveyed neither agreed nor disagreed with this statement. The results suggest,

however, that migration is perceived as a more effective preservation technique than

either emulation or data archeology. 79% of respondents agreed or strongly agreed that

migration is an effective preservation strategy for text and image files. A much lower

percentage, however, either agreed or strongly agreed that emulation and data

archeology are effective preservation strategies (12 -17% and 17-21% respectively, as

the percentage varies based on type of file)

No respondent disagreed or strongly disagreed that migration is an effective

digital preservation technique for image and text. However, 17% disagree that

 

38

 

emulation is an effective preservation strategy for digital text and images and 29 – 33%

disagree that data archeology is an effective preservation strategy for digital text and

images.

Perceptions of the popularity of Migration, Emulation, and Digital Archeology

A high percentage of respondents neither agreed nor disagreed with the

statements concerning the popularity of migration (62%), emulation (46%), and data

archeology(67%). Most participants (50%), disagreed or strongly disagreed that

emulation is widely implemented while 4% agreed or strongly agreed with this

statement. The results suggest that migration is perceived to be most widely

implemented. 29% of individuals surveyed agreed or strongly agreed that migration is

widely implemented while 8% of individuals disagreed with this statement. No

participant agreed that data archaeology is widely implemented. However, only 33 %

disagreed or strongly disagreed with this statement. Figure 12 illustrates the exact

counts of survey responses to the statement that migration, emulation, and data

archeology (respectively) are widely implemented.

Figure 12: Institutional perceptions of the statement that migration, emulation, and data archeology

(respectively) is widely implemented, by institution count.

 

39

 

Neither Agree nor Disagree

With the exception of a substantial percentage of agreement with statements

relating to migration (as discussed earlier), a large percentage of respondents neither

agreed nor disagreed with each statement made about each digital preservation strategy.

Figure 13 illustrates the number of ‘neither agree nor disagree’ responses given for all

questions relating to the perceptions of migration, emulation, and data archeology.

Figure 13: Counts of institutions that neither agree nor disagree with statements made about migration,

emulation, and data archeology.

More respondents indicated that they neither agreed nor disagreed that

emulation is an effective digital preservation strategy for images and text than those that

indicated the same opinion for migration or data archeology. Emulation, overall, has the

greatest amount of neither-agree-nor-disagree responses, except for perceptions of its

popularity. Figure 14 reveals that emulation is perceived as the least widely

implemented preservation technique of the three. These two results support the claim

that ARL institutions are more familiar with migration and data archeology than

 

40

 

emulation. Migration has the least amount of respondents that neither agreed nor

disagreed that it is an effective preservation strategy. Data archeology and migration

have similar levels of respondents that neither agreed nor disagreeed with the statements

made about them.

Figure 14: Counts of institutions that disagree or strongly disagree with statements made about migration,

emulation, and data archeology.

Participant Perceptions of Digital Preservation Techniques that their Institution has

Adopted

A major research question for this study proposes to find if there is a

relationship between the use of certain digital preservation techniques and the attitudes

these institutions display regarding these particular techniques. The results indicate that

there are less respondents that neither agree nor disagree with statements made about

the digital preservation technique among the population that has engaged in this

technique than among the general population.The degree to which this is true varies

 

41

 

based on the digital preservation technique in question. Each preservation technique is

analyzed below.

The Perceptions of participants who have participated in migration regarding the

strategy of migration

10 of 25 respondents indicated that their institution has engaged in migration of

their digital image files, of these respondents 6 also have migrated their digital text files.

The comparison of perceptions of institutions that have migrated and those that have not

show the least amount of difference in the amount of the respondents in either group

that neither agree nor disagree with statements made about migration. Respondents who

indicate that their institutions has engaged in migration have 22% less neither agree nor

disagree opinions regarding the financial expense of migration as oposed to respondents

that have not engaged in migration. A large percentage of those who have participated

in migration (80%) agree or strongly agree that it is financially expensive, compared to

55% of the general respondents who feel this way.

Figure 15 and 16 illustrate the percentage of survey responses (strongly agree,

agree, neither agree nor disagree, disagree, strongly disagree) to statements offered

regarding migration for both the specailized group that has engaged in this strategy, and

the general group of respondents.

 

42

 

Figure 15: Perceptions of migration offered by respondents from institutions that have previously

indicated that they have engaged in migration

Figure 16: Overall perceptions of migration

 

43

 

The Perceptions of participants who have participated in emulation regarding the

strategy of emulation.

Of 25 responses, two indicated that their institution has enganged in emulation

of their digital image files or digital text files.

It is interesting to note that while over 50% of the the general population neither

agreed nor disagreed with statements regarding the amount of labor, expense, and

effectiveness of emulation, the population that has used emulation either agreed or

disagreed with all of these statements. Both individuals agree that emulation is an

effective preservation strategy for digital text and images. Both agree that emulation is

financially expensive. One agrees that emulation is labor intensive, while the other

disagrees. One agrees that emulation is widely implemented, while the other neither

agrees nor disagrees.

Figure 17 and 18 illustrate the percentage of survey responses (strongly agree,

agree, neither agree nor disagree, disagree, strongly disagree) to statements offered

regarding emulation for both the specailized group that has engaged in this strategy, and

the general group of respondents.

 

44

 

Figure 17: Perceptions of emulation offered by respondents from institutions that have previously

indicated that they have engaged in emulation. N = 2.

Figure 18: General perceptions of emulation.

 

45

 

The Perceptions of participants who have participated in data archeology regarding the

strategy of data archeology.

Of 25 responses, 6 indicated that their institutiont has engaged in data

archeology for digital images and 3 for digital text. Of the 6 individuals who had used

data archeology on their digital image files, 3 had also used data archeology on their

text files. Therefore group institutions who used data archeology on image files is

inclusive of institutions who have used data archology on text files.

Interestingly, in the population of individuals who had engaged in data

archeology, there were no respondents that neither agreed nor disagreed with statements

made about intensity of labor and financial expense of this preservation technique.100%

of individuals who indicated that their institution has used data archeology agree or

strongly agree that data archeology is labor intensive and financially expensive. One-

third of respondents in the general population indicated that they neither agreed nor

disagreed with these statements, however no respondents in the general population

disagree or strongly disagree that data archeology is either labor intensive or financially

expensive.  

Figure 19 and 20 illustrate the percentage of survey responses (strongly agree,

agree, neither agree nor disagree, disagree, strongly disagree) to statements offered

regarding data archeology for both the specailized group that has engaged in this

strategy, and the general group of respondents.

 

46

 

Figure 19: Perceptions of data archeology for institutions that have previously indicated that they have

engaged in data archeology 

Figure 20: General perceptions of data archeology

 

47

 

Discussion

The purpose of this research is to provide an exploratory study of the digital

preservation activities and attitudes at ARL institutions with a special focus on the

processes and perspectives of grant-receiving institutions.

The results suggest that both grant receiving and non-grant receiving institutions

are more likely not to use migration, emulation, or data archeology than they are to use

them. This result mirrors the surveys conducted by Simpson in the UK, and Pymm and

Lloyd in Australia, both of which reported a low percentage of respondents involved in

a digital preservation strategy). However, contrary to Pymm and Lloyd’s 2005 findings

in which no library surveyed was involved in a migration program (176), a large

percentage of ARL institutions are engaged in migration.

The results suggest that there is no statistically significant relationship between

the ability of an institution to receive a national digitization or digital preservation grant

and the implementation of migration, emulation or digital archeology at that institution.

Respondents from institutions that had received national digitization or digital

preservation grants were just as likely to report that they engaged in migration,

emulation, and digital archeology as institutions that had not received such grants. This

seems inconsistent with the findings of the 2002 survey of US institutions and archival

projects conducted by Cloonan and Sanett (which reports a high percentage of

respondents that were engaged in a digital preservation strategy) as compared to the

results of Simpson’s survey of institutions in the UK and Pymm and Lloyd’s Australian

survey (which report a low percentage of respondents that were engaged in a digital

preservation strategy). The institutions and projects that took the American survey all

 

48

 

employed a member of InterPARES and the disparity between the results of the

American study and the UK and Australian study may be accounted for by the fact that

the American study included only institutions and projects with a demonstrated

commitment to digital curation. However, it appears that the ability of an institution to

receive a national digitization or digital preservation grant does not provide the same

litmus for the presence of digital preservation activities as participation in the

InterPARES Preservation Task Force

ARL institutions hold much more than plain text and image files in their

collections. Most also have audio and video files. Some even collect statistical datasets

and websites. The surveys conducted in the UK and Australia suggest that the digital

preservation practices for complex digital objects are similar to those for the simpler

ones. However, my survey suggests that ARL institutions are more likely to practice a

digital preservations strategy on digital image files than digital text files. Exactly why

this may be true is unclear. One may speculate a variety of possible reasons:

respondents may believe that digital image files as more vulnerable or valuable to the

institution than text files, the literature may have been more successful in demonstrating

the importance of digital preservation to image files, or the community may be more

active in creating standardize file formats for images. While the survey results for

digital text file formats were more varied, the results for digital image files formats

were overwhelmingly consistent. Despite Buonora and Liberati’s claim that JPEG 2000

has become the defacto standard for digital image archiving, it seems that TIFF remains

the most common digital image file format in ARL institutions.

 

49

 

Nearly 2 in 5 institutions have migrated image files, 1 in 4 have used data

archeology and 1 in 12 have emulated image files. In agreement with Kranch, Bullock,

Granger and Rothenberg, migration is the most common digital preservation strategy

used in ARL institutions. It is surprising that ARL institutions prefer data archeology

(retro-archiving) to emulation due to the perceived high cost and amount of labor

involved in digital archeology (Rothenberg). However, it seems that ARL institutions

are well aware of the intensity of the labor and high financial cost involved with digital

archeology. All of the ARL institutions, who had engaged in digital archeology

previously, acknowledged that data-archiving is both labor intensive and financially

expensive. Additionally, ARL institutions generally considered data archeology more

expensive than migration or emulation.

The results suggest that the preservation perspectives of institutions that have

engaged in various digital preservation techniques tend to differ from those which have

not. Namely there are fewer respondents that neither agree nor disagree with statements

made about specific preservation techniques that the institution has engaged in. The

trend of neutrality in the perspectives of digital preservation techniques will be

discussed later, but it may be important to prelude that discussion with the claim that

has been illustrated by the preceding result – that the level of neutrality regarding

statements made about various digital preservation techniques lessens as the participant

has more personal experience with it.

The results display a high percentage of institutional neutrality over emulation,

data archeology, and at times migration. This suggests that these techniques are not well

understood. The literature provides many possible reasons why this is a valid claim.

 

50

 

Although there is much literature available, there is also much disagreement over which

is the ‘best strategy’.

It is important to emphasize the trend of neutrality regarding digital preservation

perspectives as a large portion of institutions provided a neither agree nor disagree

response to various statements made about digital preservation strategies. While the

numbers suggest that ARL institutions consider migration more effective than

emulation, there may be a large number of institutions who have not made the final

determination, enough to sway the numbers either way. The survey offers insight into

the perceptions of ARL institutions limited to one point in time. Among ARL

institutions, data archeology was considered the most labor intensive and the most

expensive. These results comply with Rothenberg’s critique. Migration was a close

second to data archeology in terms of labor and expense. However, migration was

considered the most effective digital preservation strategy for digital image and text

files. Emulation was marked by neutrality, except for perceptions of its popularity: it is

considered the least widely implemented technique of the three.

It is also important to address the theme of neutrality in light of the individuals

who were recruited to the survey. The head administrator of the library was asked to

pass the recruitment message along to the individual that he or she considered most

capable of providing the digital preservation practices of the institution. The job titles of

these individuals support Hedstrom’s statement that digital preservation has become the

concern of more than just the archivist. The majority of job titles did not indicate

specialization in archiving or preservation, but rather indicated specialization in digital

technologies. One may speculate that the prevailing neutrality towards preservation

 

51

 

techniques may be the result of a lack of formal education in digital curation and

archiving. However, the survey did not ask for educational background, and the data

cannot test this hypothesis.

Closing Remarks

Millions of dollars are being invested in digitization and digital acquisitions at

libraries, but if these resources are not adequately preserved, they will become nearly

irrevocably obsolete and thereby inaccessible. Certainly, we cannot expect grant

funding agencies to fund repeat projects in the future to re-digitize or re-acquire what

has been mishandled, in cases when that is even possible. Additionally, library users

have grown accustomed to the convenience of digital materials, and one may expect

that digital resources will become even more demanded by library users in the future.

Adequately applying digital preservation principles to the vast quantities of digital

materials in libraries will not only help ensure their long-term survival but it will also

help libraries be more effective stewards of their resources. It is hoped that the findings

of this research will inform academic librarians and digital archivists of the practices of

academic libraries in the realm of digital preservation. Libraries may be, understandably

overwhelmed with the issues of digital preservation. Many libraries may view digital

preservation as beyond the scope of their resources. Therefore, a large priority of the

research is to identify the attitudes that libraries have towards digital preservation

techniques. Additionally, if digital preservation in archival institutions is truly as

inactive as previous surveys indicate, perhaps knowledge of common perceptions will

help experts frame their discussions in a manner that serves the academic library, and

 

52

 

remove the mystery associated with digital preservation in order to encourage the

implementation of these techniques.

Suggestions for future research

While it may be clear that many institutions world-wide are not engaged in the

digital preservation of their digital collections, the reasons why are less obvious. This

study suggests that many individuals are not familiar with digital preservation

techniques, but how does that play into library politics? McGovern reinforces the

importance of organization in digital preservation through her metaphor of the three-

legged stool. Are institutions not actively curating their digital collections because they

do not know how? Is it because the administration is not convinced of the importance?

Is it a matter of organizational support? Do they perceive their digital collections as still

being so new that they do not yet have to worry about implementing digital preservation

activities? Future research may be needed to reveal the reasons why institutions do not

involve themselves in a digital preservation program.

 

53

 

Works Cited

Babbie, Earl. The Practice of Social Research. Bellmont: Wadsworth, 2007.

Balkestein, Marjan, and Heiko Tjalsma. “The ADA Approach: Retro-archiving Data in

an Academic Environment.” Archival Science 7.1 (2007): 89-105.

Bhatnagar, Anjana. “Digitization in Academic Libraries.” Information Studies 12.1

(2006): 35-54.

Buckley, Robert. “JPEG 2000 – a Practical Digital Preservation Standard?” Digital

Preservation Coalition February 2008: 26 March 2009.

<http://www.dpconline.org/docs/reports/dpctw08-01.pdf>.

Bullock, Alison. “Preservation of Digital Information: Issues and Current Status.”

National Library of Canada. 22 April 1999. 19 October 2008.< http://epe.lac-

bac.gc.ca/100/202/301/netnotes/netnotes-h/notes60.htm>.

Buonora, Paolo, and Franco Liberati. “A Format for Digital Preservation of Images: A

Study on JPEG 2000 File Robustness.” D-Lib Magazine. 14.7/8. 21 October

2008. <http://www.dlib.org/dlib/july08/buonora/07buonora.html#1>.

Cloonan, Michele V., and Shelby Sanett. “Preservation Strategies for Electronic

Records: Where We Are Now—Obliquity and Squint?” The American Archivist

65(2002): 70-106.

Colin Webb, "Migration Trials: Migrating publications on floppy disk to CD-R."

October 1997. 19 October 2008.

<http://www.nla.gov.au/nla/staffpaper/cwebb7.html>.

 

54

 

Granger, Stewart. “Emulation as a Digital Preservation Strategy.” D-Lib Magazine.

October 2000. 8 October 2008. <http://webdoc.sub.gwdg.de/edoc/aw/d-

lib/dlib/october00/granger/10granger.html>.

Hedstrom, Margaret, and Christopher Lee “Significant Properties of Digital Objects:

Definitions, Applications, Implications.” Proceedings of the DLM-Forum 2002,

Barcelona, 6–8 May 2002. Luxembourg: Office for Official Publications of the

European Communities, 2002. 218-227.

Hedstrom, Margaret. “It's About Time: Research Challenges in Digital Archiving and

Long-term Preservation.” Washington, DC: The National Science Foundation

and the Library of Congress, Feb 2003. 18 Oct 2008.

<http://www.si.umich.edu/digarch/NSF%200915031.pdf>.

Hedstrom, Margaret, Christopher Lee, Judith Olson, and Clifford Lampe. “The Old

Version Flickers More: Digital Preservation from the User’s Perspective.” The

American Archivist 69 (2006): 159-187.

Kranch, Douglas A. “Beyond Migration: Preserving Electronic Documents with Digital

Tablets.” Information Technology and Libraries. 17.3 (1998). 20 Oct 2008.

<http://vnweb.hwwilsonweb.com/hww/results/results_single_fulltext.jhtml;hww

ilsonid=FDXCTNJHGFNY5QA3DIKSFGOADUNGIIV0>.

Mellor, Phil, Paul Wheatley and Derek Sergeant. “Migration on Request, a Practical

Technique for Preservation.” Research and Advanced Technology for Digital

Libraries. (2002). 516-526. 8 October 2008.

<http://www.springerlink.com/content/752vmvw0g0w40dj2/fulltext.pdf >\

 

55

 

Oltmans, Erik, and Nanda Kol. “A Comparison Between Migration and Emulation in

Terms of Costs.” RLG Digi-News 9.10 (2007). 8 Oct 2008.

<http://worldcat.org/arcviewer/1/OCC/2007/08/08/0000070519/viewer/file959.h

tml#article0>.

Peterson, Kit comp. “Standards Related to Digital Imaging of Pictorial Materials.”

Library of Congress. 2004. 10 Oct 2008.

<http://www.loc.gov/rr/print/tp/DigitizationStandardsPictorial.pdf >.

Pymm, Bob, and Annemaree Lloyd. “Dealing with Digital Collections: Interviews with

the National Library and Selected State Libraries of Australia.” Australian

Academic & Research Libraries. 38.3 (2007): 167 – 179.

Reference Model for an Open Archival Information System (OAIS) Blue Book. CCSDS

650.0.B-1. January 2002. 20 Oct 2008.

<http://public.ccsds.org/publications/archive/650x0b1.pdf >.

Rothenberg, Jeff. “Avoiding Technological Quicksand: Finding a Viable Technical

Foundation for Digital Preservation.” CLIR Reports. Jan 1998. 8 Oct 2008.

<http://www.clir.org/pubs/reports/rothenberg/contents.html>.

Stanescu, Andreas. “Assessing the Durability of Formats in a Digital Preservation

Environment” OCLC Systems & Services. 21.1 (2005): 61-81.

Simpson, Duncan. “Digital Preservation in the Regions Museums, Libraries and

Archives Council.” 2005. 20 Oct 2008.

<http://www.mla.gov.uk/resources/assets//M/mla_dpc_survey_pdf_6636.pdf>.

 

56

 

Appendices

Appendix A: The Survey

1. Please identify the location of your institution (U.S. Canada Other)

2. Please indicate your job title.

3. Please provide the name of your institution. (Note: this information is used to identify duplicate responses and will not be included in any derived report).

4. Within the past 10 years, has the institution been involved in a national digitization and/or digital preservation grant? (Y N Applied but didn’t receive funding Don't know)

5. Please describe the contents (e.g. text, images) and size of your institution’s digital collections.

6. Please provide the following information for digital TEXT files: a. What file format(s) do you use for the digital master copies of your digital text files?

b. On what physical media do you store your digital text collections? c. Are you storing your digital text collections on physical media that are different

from those to which you originally stored them? d. What is the age range of these materials? 7. Please provide the following information for digital IMAGE files:

a. What file format(s) do you use for the digital master copies of your digital image files?

b. On what physical media do you store your digital image collections? c. Are you storing your digital text collections on physical media that are different

from those to which you originally stored them? d. What is the age range of these materials? 8. Please indicate whether your institution has used migration in the management of

your digital image files Yes No Don’t Know 9. Please indicate whether your institution has used emulation in the management of

your digital image files Yes No Don’t Know 10. Please indicate whether your institution has used data archeology (extracting

information from obsolete media – yourself or by hiring a vendor to do the extraction) in the management of your digital image files

Yes No Don’t Know

 

57

 

 

11. Please indicate whether your institution has used migration in the management of

your digital text files Yes No Don’t Know 12. Please indicate whether your institution has used emulation in the management of

your digital text files Yes No Don’t Know 13. Please indicate whether your institution has used data archeology(extracting

information from obsolete media – yourself or by hiring a vendor to do the extraction) in the management of your digital text files

Yes No Don’t Know 14. Please indicate your level of agreement with the following statements: Migration: is labor intensive has a high financial cost is an effective preservation strategy for digital text is an effective preservation strategy for digital images is widely implemented Emulation: is labor intensive has a high financial cost is an effective preservation strategy for digital text is an effective preservation strategy for digital images is widely implemented Digital Archeology (extracting information from obsolete media – yourself or by hiring a vendor to do the extraction): is labor intensive has a high financial cost is an effective preservation strategy for digital text is an effective preservation strategy for digital images is widely implemented